From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stephen Warren <swarren@wwwdotorg.org>
Date: Fri, 14 Feb 2014 23:38:23 -0700
Subject: [U-Boot] [PATCH v2] usb: tegra: Fix PHY configuration for Tegra
	3
In-Reply-To: <4B9C9637D5087840A465BDCB251780E9E3373B1191@HKMAIL02.nvidia.com>
References: <97719287199013e3488b50b9c7e70e21f38400dd.1392417695.git.stefan@agner.ch>,
	<952d768092b4ac9912931e9e65e3adc6@agner.ch>
	<4B9C9637D5087840A465BDCB251780E9E3373B1191@HKMAIL02.nvidia.com>
Message-ID: <52FF0B5F.9000707@wwwdotorg.org>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

On 02/14/2014 09:47 PM, Jim Lin wrote:
> 
> ________________________________________
> ???: Stefan Agner [stefan at agner.ch]
> ????: 2014?2?15? ?? 06:52
> ???: u-boot at lists.denx.de; Jim Lin; Tom Warren; swarren at wwwdotorg.org; sjg at chromium.org; dev at lynxeye.de
> ??: Re: [PATCH v2] usb: tegra: Fix PHY configuration for Tegra 3
> 
> Am 2014-02-14 23:45, schrieb Stefan Agner:
>>       /* Select ULPI parallel interface */
>> -     clrsetbits_le32(&usbctlr->port_sc1, PTS_MASK, PTS_ULPI << PTS_SHIFT);
>> +     if (!controller->has_hostpc) {
>> +             clrsetbits_le32(&usbctlr->port_sc1, PTS_MASK,
>> +                             PTS_ULPI << PTS_SHIFT);
>> +     else
>> +             clrsetbits_le32(&usbctlr->hostpc1_devlc, PTS_MASK,
>> +                             PTS_ULPI << PTS_SHIFT);
>>
>>       /* enable ULPI transceiver */
>>       setbits_le32(&usbctlr->susp_ctrl, ULPI_PHY_ENB);
> 
> Ok, just noticed that we need to address the different USBD register
> layout on Tegra 2 here too.
> 
> Furthermore, the code does not compile with Tegra 2 header file since
> the hostpc1_devlc field is missing there. The function
> ehci_get_port_speed uses a define and some calculation to work around
> this issue. Another solution would be to create a dummy field in the
> Tegra 2 USB register header file... Any thoughts on that?
> 
> [Jim} Try use
>  usbctlr->usb_cmd + HOSTPC1_DEVLC 
> instead of  usbctlr->hostpc1_devlc

I don't think we should mix register access styles. Either ifdef the
driver so that only fields appropriate for the SoC being compiled for
are accessed, or if the newer SoCs are just supersets of the earlier
ones, mere the USB headers together so we don't need to split it up per
SoC (and probably add comments indicating which fields are relevant for
which SoCs). Actually, you could probably merge the headers even if you
go down the ifdef route, to avoid duplicating 99% of the header for each
SoC.